Electromembrane extraction (EME) is a miniaturized technique in which charged molecules are extracted from biological samples, through a porous membrane, into an acceptor solution. This process occurs in the presence of an electric field that enables increased analyte recovery in short extraction times. Studies are being carried out to establish recommended extraction conditions for different groups of compounds based on their physicochemical properties. In this work, pre-established generic liquid membranes for the extraction of basic compounds were tested using synthetic cathinones within a wide log P range as basic model analytes. First, recommended organic solvents for the chosen analytes were evaluated in terms of recovery. The best candidates were included in an optimization study with other relevant EME conditions (voltage, extraction time, and agitation rate). All analyses were performed using a UPLC-MS/MS with the multiple reaction monitoring mode. 2-Undecanone was chosen as the liquid membrane providing the best extraction efficiency in combination with 40 V, 35 min, and 725 rpm. Validation showed r² > 0.99 in the 1–500 ng/mL range, accuracy of ±12%, imprecision of ≤19%, 1 ng/mL as limit of quantitation, 0.1–0.5 ng/mL as limit of detection, matrix effect 97–118%, and recovery 63–110%. Among the generic liquid membranes studied in this work, 2-Undecanone was the best choice for the target basic analytes. High recoveries and low matrix effects were achieved, thus demonstrating the applicability of recommended generic liquid membranes in the 96-well plate setup and with whole blood samples. Our findings contribute to further establishing EME as an actual sample preparation technique for routine analyses that could be further explored in a commercial format.